Voice recognition, also known as (STT), is a computer processing subsystem in which a person speaks into a transducer, such as a microphone in a natural tone of voice. The transducer captures and digitizes the speech, and a speech-to-text engine parses the speech into words or phrases, which are then conditioned and compared to baseline models to create computer-readable text. Thus, in a theoretically ideal STT engine, a user could read the text of this section of the specification into a microphone, and the STT would reproduce the corresponding ASCII text verbatim.